Classifier selection using sequential error ratio criterion for multi-instance and multi-sample fusion

Classifier selection is a problem encountered by multi-biometric systems that aim to improve performance through fusion of decisions. A particular decision fusion architecture that combines multiple instances (n classifiers) and multiple samples (m attempts at each classifier) has been proposed in previous work to achieve controlled trade-off between false alarms and false rejects. Although analysis on text-dependent speaker verification has demonstrated better performance for fusion of decisions with favourable dependence compared to statistically independent decisions, the performance is not always optimal. Given a pool of instances, best performance with this architecture is obtained for certain combination of instances. Heuristic rules and diversity measures have been commonly used for classifier selection but it is shown that optimal performance is achieved for the `best combination performance' rule. As the search complexity for this rule increases exponentially with the addition of classifiers, a measure - the sequential error ratio (SER) - is proposed in this work that is specifically adapted to the characteristics of sequential fusion architecture. The proposed measure can be used to select a classifier that is most likely to produce a correct decision at each stage. Error rates for fusion of text-dependent HMM based speaker models using SER are compared with other classifier selection methodologies. SER is shown to achieve near optimal performance for sequential fusion of multiple instances with or without the use of multiple samples. The methodology applies to multiple speech utterances for telephone or internet based access control and to other systems such as multiple finger print and multiple handwriting sample based identity verification systems.

Sequential decision fusion of multibiometrics applied to text-dependent speaker verification for controlled errors

Reliability of the performance of biometric identity verification systems remains a significant challenge. Individual biometric samples of the same person (identity class) are not identical at each presentation and performance degradation arises from intra-class variability and inter-class similarity. These limitations lead to false accepts and false rejects that are dependent. It is therefore difficult to reduce the rate of one type of error without increasing the other. The focus of this dissertation is to investigate a method based on classifier fusion techniques to better control the trade-off between the verification errors using text-dependent speaker verification as the test platform. A sequential classifier fusion architecture that integrates multi-instance and multisample fusion schemes is proposed. This fusion method enables a controlled trade-off between false alarms and false rejects. For statistically independent classifier decisions, analytical expressions for each type of verification error are derived using base classifier performances. As this assumption may not be always valid, these expressions are modified to incorporate the correlation between statistically dependent decisions from clients and impostors. The architecture is empirically evaluated by applying the proposed architecture for text dependent speaker verification using the Hidden Markov Model based digit dependent speaker models in each stage with multiple attempts for each digit utterance. The trade-off between the verification errors is controlled using the parameters, number of decision stages (instances) and the number of attempts at each decision stage (samples), fine-tuned on evaluation/tune set. The statistical validation of the derived expressions for error estimates is evaluated on test data. The performance of the sequential method is further demonstrated to depend on the order of the combination of digits (instances) and the nature of repetitive attempts (samples). The false rejection and false acceptance rates for proposed fusion are estimated using the base classifier performances, the variance in correlation between classifier decisions and the sequence of classifiers with favourable dependence selected using the 'Sequential Error Ratio' criteria. The error rates are better estimated by incorporating user-dependent (such as speaker-dependent thresholds and speaker-specific digit combinations) and class-dependent (such as clientimpostor dependent favourable combinations and class-error based threshold estimation) information. The proposed architecture is desirable in most of the speaker verification applications such as remote authentication, telephone and internet shopping applications. The tuning of parameters - the number of instances and samples - serve both the security and user convenience requirements of speaker-specific verification. The architecture investigated here is applicable to verification using other biometric modalities such as handwriting, fingerprints and key strokes.

Sequential decision fusion for controlled detection errors

Information fusion in biometrics has received considerable attention. The architecture proposed here is based on the sequential integration of multi-instance and multi-sample fusion schemes. This method is analytically shown to improve the performance and allow a controlled trade-off between false alarms and false rejects when the classifier decisions are statistically independent. Equations developed for detection error rates are experimentally evaluated by considering the proposed architecture for text dependent speaker verification using HMM based digit dependent speaker models. The tuning of parameters, n classifiers and m attempts/samples, is investigated and the resultant detection error trade-off performance is evaluated on individual digits. Results show that performance improvement can be achieved even for weaker classifiers (FRR-19.6%, FAR-16.7%). The architectures investigated apply to speaker verification from spoken digit strings such as credit card numbers in telephone or VOIP or internet based applications.

Interleaved PC-OFDM to reduce peak-to-average power ratio

Parallel combinatory orthogonal frequency division multiplexing (PC-OFDM yields lower maximum peak-to-average power ratio (PAR), high bandwidth efficiency and lower bit error rate (BER) on Gaussian channels compared to OFDM systems. However, PC-OFDM does not improve the statistics of PAR significantly. In this chapter, the use of a set of fixed permutations to improve the statistics of the PAR of a PC-OFDM signal is presented. For this technique, interleavers are used to produce K-1 permuted sequences from the same information sequence. The sequence with the lowest PAR, among K sequences is chosen for the transmission. The PAR of a PC-OFDM signal can be further reduced by 3-4 dB by this technique. Mathematical expressions for the complementary cumulative density function (CCDF)of PAR of PC-OFDM signal and interleaved PC-OFDM signal are also presented.

Cross likelihood ratio based speaker clustering using eigenvoice models

This paper proposes the use of eigenvoice modeling techniques with the Cross Likelihood Ratio (CLR) as a criterion for speaker clustering within a speaker diarization system. The CLR has previously been shown to be a robust decision criterion for speaker clustering using Gaussian Mixture Models. Recently, eigenvoice modeling techniques have become increasingly popular, due to its ability to adequately represent a speaker based on sparse training data, as well as an improved capture of differences in speaker characteristics. This paper hence proposes that it would be beneficial to capitalize on the advantages of eigenvoice modeling in a CLR framework. Results obtained on the 2002 Rich Transcription (RT-02) Evaluation dataset show an improved clustering performance, resulting in a 35.1% relative improvement in the overall Diarization Error Rate (DER) compared to the baseline system.

Sequential Fusion Using Correlated Decisions for Controlled Verification Errors

Fusion techniques have received considerable attention for achieving lower error rates with biometrics. A fused classifier architecture based on sequential integration of multi-instance and multi-sample fusion schemes allows controlled trade-off between false alarms and false rejects. Expressions for each type of error for the fused system have previously been derived for the case of statistically independent classifier decisions. It is shown in this paper that the performance of this architecture can be improved by modelling the correlation between classifier decisions. Correlation modelling also enables better tuning of fusion model parameters, ‘N’, the number of classifiers and ‘M’, the number of attempts/samples, and facilitates the determination of error bounds for false rejects and false accepts for each specific user. Error trade-off performance of the architecture is evaluated using HMM based speaker verification on utterances of individual digits. Results show that performance is improved for the case of favourable correlated decisions. The architecture investigated here is directly applicable to speaker verification from spoken digit strings such as credit card numbers in telephone or voice over internet protocol based applications. It is also applicable to other biometric modalities such as finger prints and handwriting samples.

Sequential fusion of decisions from adaptive and random samples for controlled verification of errors

Fusion techniques have received considerable attention for achieving performance improvement with biometrics. While a multi-sample fusion architecture reduces false rejects, it also increases false accepts. This impact on performance also depends on the nature of subsequent attempts, i.e., random or adaptive. Expressions for error rates are presented and experimentally evaluated in this work by considering the multi-sample fusion architecture for text-dependent speaker verification using HMM based digit dependent speaker models. Analysis incorporating correlation modeling demonstrates that the use of adaptive samples improves overall fusion performance compared to randomly repeated samples. For a text dependent speaker verification system using digit strings, sequential decision fusion of seven instances with three random samples is shown to reduce the overall error of the verification system by 26% which can be further reduced by 6% for adaptive samples. This analysis novel in its treatment of random and adaptive multiple presentations within a sequential fused decision architecture, is also applicable to other biometric modalities such as finger prints and handwriting samples.

Sequential fusion of decisions with favorable dependence for controlled verification errors

Statistical dependence between classifier decisions is often shown to improve performance over statistically independent decisions. Though the solution for favourable dependence between two classifier decisions has been derived, the theoretical analysis for the general case of 'n' client and impostor decision fusion has not been presented before. This paper presents the expressions developed for favourable dependence of multi-instance and multi-sample fusion schemes that employ 'AND' and 'OR' rules. The expressions are experimentally evaluated by considering the proposed architecture for text-dependent speaker verification using HMM based digit dependent speaker models. The improvement in fusion performance is found to be higher when digit combinations with favourable client and impostor decisions are used for speaker verification. The total error rate of 20% for fusion of independent decisions is reduced to 2.1% for fusion of decisions that are favourable for both client and impostors. The expressions developed here are also applicable to other biometric modalities, such as finger prints and handwriting samples, for reliable identity verification.

On the diversity analysis of the sequential slotted amplify-decode-and-forward protocol

In this paper, we review the sequential slotted amplify-decode-and-forward (SADF) protocol with half-duplex single-antenna and evaluate its performance in terms of pairwise error probability (PEP). We obtain the PEP upper bound of the protocol and find out that the achievable diversity order of the protocol is two with arbitrary number of relay terminals. To achieve the maximum achievable diversity order, we propose a simple precoder that is easy to implement with any number of relay terminals and transmission slots. Simulation results show that the proposed precoder achieves the maximum achievable diversity order and has similar BER performance compared to some of the existing precoders.

Sequential slotted amplify-decode-and-forward

In this paper, we propose a novel slotted hybrid cooperative protocol named the sequential slotted amplify-decodeand-forward (SADF) protocol and evaluate its performance in terms of diversity-multiplexing trade-off (DMT). The relays between the source and destination are divided into two different groups and each relay either amplifies or decodes the received signal. We first compute the optimal DMT of the proposed protocol with the assumption of perfect decoding at the DF relays. We then derive the DMT closed-form expression of the proposed sequential-SADF and obtain the proximity gain bound for achieving the optimal DMT. With the proximity gain bound, we then found the distance ratio to achieve the optimal DMT performance. Simulation result shows that the proposed protocol with high proximity gain outperforms other cooperative communication protocols in high SNR regime.

Eigenvoice modeling for cross likelihood ratio based speaker clustering : a Bayesian approach

This paper proposes the use of Bayesian approaches with the cross likelihood ratio (CLR) as a criterion for speaker clustering within a speaker diarization system, using eigenvoice modeling techniques. The CLR has previously been shown to be an effective decision criterion for speaker clustering using Gaussian mixture models. Recently, eigenvoice modeling has become an increasingly popular technique, due to its ability to adequately represent a speaker based on sparse training data, as well as to provide an improved capture of differences in speaker characteristics. The integration of eigenvoice modeling into the CLR framework to capitalize on the advantage of both techniques has also been shown to be beneficial for the speaker clustering task. Building on that success, this paper proposes the use of Bayesian methods to compute the conditional probabilities in computing the CLR, thus effectively combining the eigenvoice-CLR framework with the advantages of a Bayesian approach to the diarization problem. Results obtained on the 2002 Rich Transcription (RT-02) Evaluation dataset show an improved clustering performance, resulting in a 33.5% relative improvement in the overall Diarization Error Rate (DER) compared to the baseline system.

Novel sequential methods in dose-finding designs : extensions of the continual reassessment method

Dose-finding trials are a form of clinical data collection process in which the primary objective is to estimate an optimum dose of an investigational new drug when given to a patient. This thesis develops and explores three novel dose-finding design methodologies. All design methodologies presented in this thesis are pragmatic. They use statistical models, incorporate clinicians' prior knowledge efficiently, and prematurely stop a trial for safety or futility reasons. Designing actual dose-finding trials using these methodologies will minimize practical difficulties, improve efficiency of dose estimation, be flexible to stop early and reduce possible patient discomfort or harm.

Sequential monitoring of hospital adverse events when control charts fail : the example of fall injuries in hospitals

Objective To evaluate methods for monitoring monthly aggregated hospital adverse event data that display clustering, non-linear trends and possible autocorrelation. Design Retrospective audit. Setting The Northern Hospital, Melbourne, Australia. Participants 171,059 patients admitted between January 2001 and December 2006. Measurements The analysis is illustrated with 72 months of patient fall injury data using a modified Shewhart U control chart, and charts derived from a quasi-Poisson generalised linear model (GLM) and a generalised additive mixed model (GAMM) that included an approximate upper control limit. Results The data were overdispersed and displayed a downward trend and possible autocorrelation. The downward trend was followed by a predictable period after December 2003. The GLM-estimated incidence rate ratio was 0.98 (95% CI 0.98 to 0.99) per month. The GAMM-fitted count fell from 12.67 (95% CI 10.05 to 15.97) in January 2001 to 5.23 (95% CI 3.82 to 7.15) in December 2006 (p<0.001). The corresponding values for the GLM were 11.9 and 3.94. Residual plots suggested that the GLM underestimated the rate at the beginning and end of the series and overestimated it in the middle. The data suggested a more rapid rate fall before 2004 and a steady state thereafter, a pattern reflected in the GAMM chart. The approximate upper two-sigma equivalent control limit in the GLM and GAMM charts identified 2 months that showed possible special-cause variation. Conclusion Charts based on GAMM analysis are a suitable alternative to Shewhart U control charts with these data.

Early stopping by using stochastic curtailment in a three-arm sequential trial

Summary. Interim analysis is important in a large clinical trial for ethical and cost considerations. Sometimes, an interim analysis needs to be performed at an earlier than planned time point. In that case, methods using stochastic curtailment are useful in examining the data for early stopping while controlling the inflation of type I and type II errors. We consider a three-arm randomized study of treatments to reduce perioperative blood loss following major surgery. Owing to slow accrual, an unplanned interim analysis was required by the study team to determine whether the study should be continued. We distinguish two different cases: when all treatments are under direct comparison and when one of the treatments is a control. We used simulations to study the operating characteristics of five different stochastic curtailment methods. We also considered the influence of timing of the interim analyses on the type I error and power of the test. We found that the type I error and power between the different methods can be quite different. The analysis for the perioperative blood loss trial was carried out at approximately a quarter of the planned sample size. We found that there is little evidence that the active treatments are better than a placebo and recommended closure of the trial.

Error bounds for calculation of the Gittins indices

For a wide class of semi-Markov decision processes the optimal policies are expressible in terms of the Gittins indices, which have been found useful in sequential clinical trials and pharmaceutical research planning. In general, the indices can be approximated via calibration based on dynamic programming of finite horizon. This paper provides some results on the accuracy of such approximations, and, in particular, gives the error bounds for some well known processes (Bernoulli reward processes, normal reward processes and exponential target processes).